feat(orchestration): Harden worker lifecycle and cleanup gates - #2
Merged
Merged
Conversation
Add durable worker manifests, deterministic state and Git/review gates, verified Herdr operations, bounded dispatch policy, and dry-run-first cleanup. Add offline validation, fake command fixtures, and G1-G10 golden scenarios so completion and teardown claims are evidence-based. Co-Authored-By: Claude <noreply@anthropic.com>
Wire the state machine, Git/review/check gate, and worker-report validator into complete and reconcile. Add run-ID manifest resolution, file-safe manifest storage, and end-to-end CLI golden coverage for completion and cleanup. Co-Authored-By: Claude <noreply@anthropic.com>
Run skill validation, offline unit tests, golden scenarios, compilation, and whitespace checks on pushes and pull requests. Co-Authored-By: Claude <noreply@anthropic.com>
Keep goal metadata and Python bytecode out of the feature worktree status without changing the main checkout. Co-Authored-By: Claude <noreply@anthropic.com>
Add one shared, explicit-opt-in Hax backend with validated configuration, safe preflight, argv construction, one-shot classification, manifest capabilities, and backend-aware dispatch limits. Integrate thin Herdr, tmux, and WezTerm adapters while preserving Pi defaults and common lifecycle gates. Add deterministic fake-command coverage, independent H1-H13 golden scenarios, runtime references, diagnostics, examples, and CI execution. Never expose credentials or silently fall back from Hax to Pi. Co-Authored-By: Claude <noreply@anthropic.com>
Make the shared Hax backend own interactive start, readiness, state, send, interrupt, resume, and stop semantics. Keep Herdr, tmux, and WezTerm adapters focused on pane transport and exact owned-target cleanup diagnostics. Add fake transport lifecycle and process-stop coverage while preserving Pi defaults and the existing H1-H13 gate. Co-Authored-By: Claude <noreply@anthropic.com>
Include backend capabilities and limitations in tmux and WezTerm status output, add exact per-runtime Hax status tests, and make CI discover those test suites. Restrict ShellCheck validation to shell files while keeping Python runtimes on Ruff and compileall. Co-Authored-By: Claude <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Turn the orchestration skills into an evidence-gated worker operations toolkit.
The change adds durable manifests, a pure lifecycle state machine, atomic redacted writes, a reliable Herdr adapter, Git/PR/review/check gates, and dry-run-first cleanup with process ownership checks. It also adds bounded dispatch controls, F15–F24 failure taxonomy coverage, offline fake-command tests, and deterministic G1–G10 golden scenarios.
Previously, the wrappers could report successful sends, setup races, idle workers, rate-limited reviews, and cleanup without durable evidence. The new CLI contracts make those conditions explicit and fail closed. Fut and tmux documentation now state their native capability limits rather than claiming unsupported state.
The main review areas are the completion/reconcile CLI integration, cleanup transaction and ownership guards, and the disposable-repository golden scenarios. No live Herdr or GitHub account is needed to run the offline gates.